I’m just gonna say it straight: after spending a stupid amount of time trying to get stock CRM lead conversion workflows working again, I do not think Corteza feels production ready in the way a serious business platform should.
And no, I’m not saying that because “one thing broke.” I’m saying it because the deeper I got into it, the more it felt like a bunch of important platform basics are either half-baked, fragile, or pushed onto the user to manually sort out. The whole thing starts to feel less like low-code and more like “hope you enjoy debugging our internals yourself.”
The first giant red flag is the promotion / migration story. If you have dev, test, and prod environments, there still does not seem to be a clean normal grown up way to move changes between them. Corteza’s own docs say namespace export/import excludes workflow references and access control permissions, which is already bad enough because that means exported apps are not really self contained. Then on the forum there are multiple threads where people are basically asking how the hell they are supposed to move changes from test to production, and Corteza staff/users are talking about DB dumps because there isn’t some proper built in app promotion flow. One user even said this is a blocker/show stopper if there’s no way to update module/page/workflow/namespace without losing productive user data. Another thread from January 2025 has someone saying they tried the DB dump route and ended up with duplicate applications and no namespaces. That’s not a real deployment story. That’s a workaround pile.
Then there’s workflows, which honestly feel way touchier than they should. The docs literally say that once you use a Prompt or Delay step, the workflow becomes asynchronous and can no longer invalidate create/update/delete actions. They also say stuff like “confirm record creation” is currently not possible with workflows. That’s a pretty major limitation for a system that’s supposed to handle business logic. So you start with a “visual workflow builder” and then quickly find out there are big caveats around what it can actually do in a clean transactional way.
And it’s not just a theoretical limitation either, because there’s enough history in their own changelogs to show workflow/prompt behavior has had real rough edges. They’ve had to fix things like bad prompt handling, prompt issues across tabs/instances, broken old-record references, and other workflow-related quirks over time. That doesn’t automatically mean current Corteza is broken everywhere, but it absolutely adds to the feeling that this area has been fragile for a while.
On top of that, some of the most frustrating workflow behavior is not obvious in the UI at all. The forum thread about workflows on selected records is a perfect example. Someone had the exact kind of greyed-out record-selection workflow issue we hit, and the answer ended up being trigger/scope related. The selected value is just a list of IDs, not the full records, and the trigger/resource setup matters more than it should for something that looks like a basic list-action workflow. That is the kind of thing the product should diagnose clearly instead of making users discover it by trial and error.
What makes this even more annoying is that I was not building some weird custom monster app. I was trying to recover stock CRM behavior. And that opens another problem: Corteza got even barer out of the box. In the 2024.9 changelog they explicitly say they removed the CRM and Service Solution apps from the default deployment. So when people say Corteza feels bare now, that’s not just vibes. They literally removed major out the box apps from the default deployment.
And once you start going beyond the most basic stuff, the “low code” label starts feeling a bit generous. Corteza’s docs say automation scripts are disabled by default since workflows were introduced, and if you need custom business logic outside workflow limits, you’re into automation scripts/Corredor territory. That may be fine for some teams, but it absolutely means parts of the platform are not really “contained” inside the visual builder the way people might expect. If you thought you were getting something where most serious logic could stay in the no-code/low-code layer, that impression fades fast.
There are also platform capability gaps that matter if you want to treat this like something enterprise-ish. Their own security model docs say access control for specific records is supported, but access control for particular fields of particular records is not. So if you need truly granular per-record per-field control, that’s a limitation right there. There are workarounds, sure, but that’s exactly the point: a lot of Corteza seems to drift into “there’s a workaround” territory instead of “this is solid and first-class.”
Even some bigger-looking platform features still come with warning labels. Federation is still marked experimental and disabled by default. Again, not the end of the world by itself, but it adds to the same overall feeling: a lot of the platform looks broad on paper, but once you dig in, parts of it still feel like “use carefully.”
And then there’s the part that really makes me question how ready this is for real business use: even stock CRM workflow recovery has already shown up on the forum as a mess. There’s a thread where someone had the CRM lead conversion buttons disabled after restore/upgrade, and the way they got it working again was exporting the workflow list from a clean install, importing it into the broken install, and then deleting/re-adding the automation buttons. That is exactly the kind of nonsense that should not be happening with out the box functionality. If stock CRM behavior can turn into restore/import/rebind/manual repair nonsense, that’s not a polished platform.
Even the import/export system has its own edge-case landmines beyond the obvious workflow-reference problem. Corteza’s changelog explicitly mentions namespace import/export edge cases that can prevent export, like pages referencing deleted modules. So even the “take your namespace and move it” path has sharp edges built into it.
So yeah, my honest opinion right now is this:
Corteza looks promising from a distance. On paper it looks flexible. In screenshots it looks like you can build a lot. But once you actually try to use it like a real platform with real environments, real workflows, real migration needs, and real expectations of stock app stability, it starts feeling way less mature than it should.
It feels bare.
It feels inconsistent.
It feels too easy to fall into undocumented or badly surfaced workflow/scope problems.
It feels like too many important things still end in “manual fix,” “DB dump it,” “re-add the buttons,” “use scripts,” or “that feature is experimental.”
That’s not what I’d call production-ready.
What I’d want to see before I could take it more seriously:
a real supported dev/test/prod promotion story
app migration that actually preserves workflow linkage and permissions
better diagnostics when workflows are disabled or missing scope data
better out the box apps instead of removing them from default deployment
less “go use scripts/Corredor” for stuff people expect the platform to handle cleanly
and way less manual repair for stock CRM behavior after restore/import
Because right now, way too much of the experience feels like the platform gives you 70% of something and expects you to fight for the last 30%. It sucks, plain and simple and needs more work before ever being implemented in a production stack.